How to Use AI Mock Vivas for MRCPCH Clinical Prep

AI mock vivas help only when they force you to sound like an MRCPCH candidate under time pressure. On Viva Examiner, use them to rehearse focused histories, parent-friendly explanations, concise sign summaries and safe management plans against the current MRCPCH Clinical Examination station types, not as a fancy flashcard app.

As of 4 July 2026, the live exam is still the face-to-face 10-station circuit with four short clinical stations, one development station, one history station, two communication stations and two video stations. RCPCH is reviewing the exam and has piloted a proposed revised format, but that is not yet the live examination; RCPCH said in March 2026 that a new MRCPCH Clinical exam is likely in 2027 at the earliest.

Why AI mock vivas help in this exam

Generic AI practice produces generic answers. This exam rewards organisation, focused data gathering, accurate information sharing, identification of signs, clinical reasoning, management planning and patient-centred communication. Those are the same domains the official station guides and anchor statements describe, so your rehearsal should be domain-based, not topic-based.

If a case goes badly, examiners rarely think that you did not read enough. More often, they hear an unfocused history, a vague summary, poor prioritisation or a management plan that is not safe enough. AI is useful because it can make you do the boring, high-yield bit: saying the answer out loud, to time, again and again.

Build an AI mock viva workflow around the MRCPCH Clinical Examination

Match each AI session to a real station. In the current circuit, history and development are 22-minute stations; communication, video and each short clinical station are 9 minutes. The four short clinical stations use real patients, the development station is scenario-based with cues, and the video stations give 3 minutes to watch the clip before examiner questioning.

Use AI differently for each station:

History: full role-play, then examiner questioning.
Communication: full role-play with emotion, concerns and interruptions.
Development: cue-based rehearsal of how you would assess development step by step.
Video: reasoning and initial management after a clip or observed-sign stem.
Short clinical: verbal summary, differential and management after you have examined a real child or practised examination technique separately.

History station: train focus, then judgement

The history station is designed to test a focused history, not a full clerking. In the live exam, you have 13 minutes for the history and 9 minutes to discuss the problems and management with the examiner; the communication domains are judged across the station as well as your information gathering and management.

A simple rehearsal structure works well:

open with your role and agenda
identify the presenting problem, timeline and severity
ask the red-flag questions that change urgency
cover relevant background only if it changes your differential or management
ask what the parent or young person is worried about
give a crisp summary
state your leading diagnosis, key differentials and first steps

A useful AI prompt is to ask the tool to act as the parent of a child with one current problem, answer only what is asked, stop the history at 13 minutes, then switch to examiner mode and press you on differential diagnosis, investigations, management and safety netting.

For example, practise with a 7-year-old with recurrent headaches. A strong answer sounds like this: you clarify red flags early, summarise in one breath, then say, 'My leading diagnosis is migraine, but I would still want to exclude raised intracranial pressure because of X and Y. My immediate next steps are...' That sounds like clinical reasoning, not guesswork.

Communication stations: explain less, connect more

RCPCH describes six broad communication patterns, including information giving, breaking bad news, consent, critical incident, ethics and education. These stations are 9 minutes long, the examiner observes throughout, and they are explicitly not a test of how much information you can pour out.

So make your AI mock viva mark you on three things:

Did you find out what the parent, adolescent or colleague already knows?
Did you give accurate information in plain English?
Did you acknowledge emotion and agree next steps?

Try this framework: open, assess starting point, explain in small chunks, check understanding, address concerns, then close with a plan. If the AI says the parent is angry after a medication error, do not launch into policy. Start with acknowledgement, apology if appropriate within your role, immediate safety and what will happen next.

Development station: make the AI give you cues

This is where AI can be genuinely useful. The live development station combines a focused developmental history, a described developmental assessment and then interpretation and management; official guidance says the examiner gives universal and dependent cues, and candidates are advised to pause briefly after each element so the cue can be delivered.

Build your prompt so the AI behaves like that examiner. Ask it to:

give the stem and any universal cues at the start
stay in role as the parent for the history phase
switch to examiner mode for the assessment description
provide a dependent cue only when you mention the relevant assessment element
challenge your interpretation and MDT plan at the end

A common failure here is sounding fluent but vague. Do not say, 'I would assess speech and language.' Say what you would actually do: how you would engage the child, what task you would offer, what milestone you expect, and what the finding would mean. Then pause.

Video and short clinical stations: use AI for the verbal half

The video stations are 9 minutes, with 3 minutes to watch the clip, 3 minutes on signs and reasoning, and 3 minutes on management. After the first 3 minutes, you cannot keep rewatching the clip, so AI is best used for the second half: describe the signs you saw, commit to the likely diagnosis, then prioritise immediate management.

The short clinical stations are different. They are patient-based stations, and the official guidance stresses focused, systematic examination technique plus interpretation and management. AI cannot replace putting your hands on children, adapting your examination to age and comfort, or learning what normal feels like. Use it after bedside practice: examine a child on the ward or in teaching, then ask AI to play examiner and ask for key signs, likely diagnosis, complications, investigations and management.

For both station types, practise short, registrar-level phrasing:

'The key signs I identified are...'
'My leading diagnosis is..., with... as an important differential.'
'My immediate priority is to assess stability, escalate if needed, and start...'

Score yourself like an examiner

RCPCH publishes anchor statements and mark sheets that use domain ratings such as meets standard, borderline and below standard. That gives you a much better debrief structure than a vague feeling that the station was good or bad.

After each AI mock viva, do a four-part debrief:

Missed data: What one question, sign or concern did I fail to uncover?
Reasoning: Did I commit to a likely diagnosis early enough?
Safety: Was my management plan specific, safe and appropriately escalated?
Communication: Did I check understanding and answer the concern that was actually in the room?

Then repeat the same case once. Not a new case. The second attempt is where you usually fix the habit that is costing marks.

Common mistakes

Using AI to test factual recall instead of spoken exam performance.
Letting the model reward long, rambling answers.
Practising short clinical stations without any real bedside examination.
Forgetting that history stations need a focused history, not a full paediatric admission clerking.
Treating communication stations as mini-lectures rather than conversations.
Failing to pause in the development station, so you never receive or process the cue properly.
Listing every possible differential before giving the most likely one.
Giving a management plan with no prioritisation, no safety net and no escalation.
Rehearsing only rare zebras instead of common paediatric problems done well.

Practice workflow on Viva Examiner

A good AI mock viva session is short and deliberate. Use a blank sheet of paper and a pen if that helps you think; note-taking is allowed in the face-to-face MRCPCH Clinical exam, although candidates leave their notes behind at the centre. RCPCH also provides official station guides, anchor statements and exam-preparation resources, including online learning and an in-person preparation course, which are useful for building realistic stems and calibration.

Try this weekly loop:

Twice a week: one history station and one communication station with full timed role-play.
Once a week: one development station using cue-based prompts.
Once a week: one video reasoning drill from a teaching clip or observed-sign description.
Twice a week: real bedside examination practice for short clinical technique.
Once a week: a human mock with a colleague who can interrupt, challenge and debrief.

Keep each AI session to one station, one objective and one fix. If today's problem is vague management planning, do not spend the session on obscure diagnoses. Spend it on saying, clearly and safely, what you would do next.

Summary

Match AI practice to the current MRCPCH Clinical Examination circuit, not a generic viva.
Use AI heavily for history, communication and development; use it as an add-on, not a substitute, for short clinical examination practice.
Score every rehearsal against domains: data gathering, reasoning, management and communication.
Repeat the same case after debrief so the correction sticks.
The best AI mock viva is the one that makes you sound clearer, safer and more organised tomorrow than you did today.

How to Use AI Mock Vivas for MRCPCH Clinical Prep

Why AI mock vivas help in this exam

Build an AI mock viva workflow around the MRCPCH Clinical Examination

History station: train focus, then judgement

Communication stations: explain less, connect more

Development station: make the AI give you cues

Video and short clinical stations: use AI for the verbal half

Score yourself like an examiner

Common mistakes

Practice workflow on Viva Examiner

Summary

References

Practice Pediatrics with a live AI examiner

Why AI mock vivas help in this exam

Build an AI mock viva workflow around the MRCPCH Clinical Examination

History station: train focus, then judgement

Communication stations: explain less, connect more

Development station: make the AI give you cues

Video and short clinical stations: use AI for the verbal half

Score yourself like an examiner

Common mistakes

Practice workflow on Viva Examiner

Summary

References

Related guides

Practice Pediatrics with a live AI examiner